The heparan sulfates of Swiss mouse 3T3 cells. The effect of transformation

Three major pools of heparan sulfate have been isolated from cultures of Swiss mouse 3T3 and SV40-transformed 3T3 cells: cell-surface, medium, and intracellular heparan sulfates. The cell-surface heparan sulfate is a high molecular weight proteogylcan which is partially degraded by pronase. Before pronase treatment, it has a peak molecular weight (as estimated by gel filtration) of appox. 7.2 · 10⁵ in contrast to only 2.4 · 10⁵ after pronase treatment. The medium heparan sulfate appears to be similar in structure to the cell-surface heparan sulfate, since they coelute on Bio-Gel A-15m and DEAE-cellulose, and are both proteoglycans. In contrast, the intracellular heparan sulfate has a low molecular weight (6.0 · 10³) and has little if any attached protein. Both the medium and intracellular heparan sulfate exhibit the transformation-associated change in structure reported earlier for cell-surface heparan sulfate (Underhill, C.B. and Keller, J.M. (1975) Biochem. Biophys. Res. Commun. 63, 448–454). This transformation-associated change, detected by DEAE-cellulose chromatography is not the result of changes in either molecular weight or protein core. Cellulose acetate electrophoresis of the cell-surface heparan sulfate at pH 1 suggests that the transformation-associated change in structure is due to a difference in sulfate content. Both types of heparan sulfate are produced in mixed cultures ot 3T3 and SV3T3 cells, indicating that neither serum factors in the culture medium nor secreted cell products are responsible for the transformation-associated change in heparan sulfate structure. The presented date are discussed with respect to the postulated role of heparan sulfate in cell social behavior.     

The effect of beta-xylosides on heparan sulfate synthesis by SV40-transformed Swiss mouse 3T3 cells.

The medium and cell surface heparan sulfates isolated from SV40-transformed Swiss mouse 3T3 cells were examined in the presence and absence of 1.0 mM p-nitrophenyl-beta-D-xyloside. Incubation of the SV3T3 cells with this beta-xyloside resulted in: (a) a 4- to 5-fold reduction in the molecular weight distribution of medium heparan sulfate, (b) a 10-fold increase in the total synthesis of medium heparan sulfate, and (c) a small reduction in cell growth. There was little, if any, change in either the total level of synthesis or the molecular weight distribution of cell surface heparan sulfate. The covalent association of the beta-xyloside to the medium heparan sulfate was demonstrated by an analysis of the medium heparan sulfate produced by cells grown in the presence of [35S]sulfate and the fluorogenic beta-xyloside, 4-methylumbelliferyl-beta-D-xyloside. Treatment of the purified radiolabeled and fluorogenic heparan sulfate with either nitrous acid or heparitinase resulted in a decrease in the molecular weight of both radiolabeled and fluorogenic material. The data presented in this paper are discussed with respect to both the structure of heparan sulfate and the putative role of heparan sulfate in cell social behavior.     

Heparan sulfates of mouse cells. Analysis of parent and transformed 3T3 cell lines.

Heparan sulfate from the surface of a variety of mouse cells at different cell densities was examined by ion-exchange chromatography. The results of this analysis show that: (1) The heparan sulfate from new isolates of Swiss 3T3 cells transformed by SV40 virus (a DNA tumor virus) elutes from DEAE-cellulose at a lower ionic strength than that from the parent cell type. This finding confirms our earlier observation with an established SV40-transformed cell line (Underhill and Keller, '75) and eliminates the possibility that this change is caused by extended passage in culture. (2) For both parent and transformed 3T3 cells, the heparan sulfates from low and high density cultures were the same as judged by chromatography on DEAE-cellulose. This result demonstrates that the transformation-dependent change which we have observed is independent of cell density. (3) The heparan sulfate from Balb/c 3T3 cells transformed with Kirsten murine sarcoma virus (an RNA tumor virus) elutes from DEAE-cellulose prior to that from parent Balb/c 3T3 cells. This result extends the transformation dependent change in heparan sulfate to the Balb/c 3T3 cell line and to cells transformed with an RNA virus.     

High molecular-weight heparan sulfate from the cell surface

Heparan sulfate fragments with molecular weight of 135,000 (as determined by equilibrium sedimentation analysis) were isolated from the trypsinate of Chinese hamster cells (line CHO) grown in culture. Evidence is presented which suggests that the intracellular heparan sulfate species with molecular weight of 10,000 to 20,000 were degradation products of the larger species. We propose that the native cell-surface heparan sulfate, in its physiological location, could serve as a nonspecific “screen” to the exposure of specific, topographically adjacent, cell-surface sites.     

The decreased synthesis of chondroitin sulfate-containing extracellular proteoglycans by SV40 transformed Balb/c 3T3 cells.

Balb/c 3T3 cells synthesize 5--10 times more 35SO2/4- -labeled extracellular proteoglycan per cell than do Balb/c 3T3 cells transformed by SV40 (SV3T3). The extracellular 35SO2/4- -labeled proteoglycans of the Balb/c 3T3 and SV3T3 cells differ markedly in their acid mucopolysaccharide composition. Extracellular Balb/c 3T3 proteoglycans contain about 70--80% chondroitin sulfate, most of which is chondroitin 4-sulfate, and small amounts of heparan sulfate and/or heparin. On the other hand, extracellular SV3T3 proteoglycans contain 65-75% heparan sulfate and/or heparin and less than 15% chondroitin sulfate. Analysis of extracellular 35SO2/4- -labeled proteoglycan by sodium dodecyl sulfate-polyacrylamide gel electrophoresis reveals that Balb/c 3T3 alone synthesizes a class of proteoglycans capable of migrating in a 10% separating gel. This class of proteoglycans, designated as fraction C, accounts for up to 45% of the total extracellular Balb/c 3T3 35 SO2/4- -labeled proteoglycans and contains chondroitin sulfate extracellular SV3T3 proteoglycans. The absence of this and other classes of chondroitin sulfate-containing proteoglycans can account for the 5-10-fold decreased synthesis of 35SO2/4- -labeled proteoglycans by SV3T3 cells when compared to Balb/c 3T3 cells.     

Altered metabolism of heparan sulfate in simian virus 40 transformed cloned mouse cells.

Glycoconjugates have been analyzed from a family of closely related mouse cells: a parent clone and three daughter subclones, two of which expressed the simian virus 40 (SV40) T-antigen. The experimental procedure involved the simultaneous comparison by DEAE-cellulose chromatography of papain-digested macromolecules from the parent, labeled with [3H]glucosamine, and one of the daughter subclones, labeled with [14C]-glucosamine. Three cultures compartments (the medium, the cell surface trypsinate, and the cells) from the paired cell lines were combined at the earliest time during the harvesting of the cells. Heparan sulfate on the surface of cells and secreted into the medium from T-antigen-positive subclones was eluted at lower salt concentrations from the anion exchange column than that from the parent clone. In the viable trypsinized cells a marked reduction of heparan sulfate was detected in the T-antigen-positive subclones. These changes were highly reproducible, were observed during both logarithmic and stationary phase of growth, and neither change was observed in the T-antigen-negative sister subclone. The elution point of heparan sulfate from Sepharose 6B was unaltered. Ratios of 35S to 3H for heparan sulfate obtained from cells doubly labeled with [35S]sulfate and [3H]glucosamine were lower in the T-antigen-positive subclones. Similar changes for the 35S to 3H ratio of chondroitin sulfate were associated with only small alterations in elution from anion exchange columns. Kinetic experiments suggested a reduced rate of incorporation of [35S]sulfate with no change in turnover rate. A substantial portion of the labeled heparan sulfate was associated with the cell surface; in contrast most of the hyaluronic acid and a large proportion of the chondroitin sulfate was apparently secreted. Quantitative changes in hyaluronic acid labeling did not correlate with expression of T-antigen. Glycosaminoglycans left on the dish after detaching cells with ethylene glycol bis(beta-aminoethyl ether)-N,N'-tetraacetic acid were nearly completely released by subsequent trypsinization. Cell detachment by trypsinization left an insignificant amount of labeled glycosaminoglycan on the dish surface. The alterations in heparan sulfate metabolism correlated with the expression of T-antigen and with the cells' ability to grow to high densities in monolayer culture, but not with growth in suspension in viscous medium. Tumorigenicity of the subclones was essentially the same as that of the parent clone.     

Heparin releases heparan sulfate from the cell surface.

Heparan sulfate of the cell surface of cultured Chinese hamster cells (line CHO) was promptly released when the cells were incubated with balanced salt solutions containing heparin. The released heparan sulfate included multichain proteoglycan of high molecular weight. The data suggest that the cell-surface localization of heparan sulfate is dependent, at least in part, upon cell-surface receptors with binding sites for the sugar chain moieties of sulfated glycosaminoglycans.     

Heparan sulfates from Swiss mouse 3T3 and SV3T3 cells: O-sulfate difference

A difference in the extent of sulfation between the heparan sulfate isolated from Swiss 3T3 mouse cells and that from Swiss 3T3 cells transformed by the DNA virus SV40 has been reported previously. This variance is manifested by different chromatographic and electrophoretic properties. Heparan sulfates from the two cell types were treated with nitrous acid under conditions that gave selective deaminative cleavage of glucosaminyl residues with sulfated amino groups in order to define the nature of the difference in sulfation further. The O-sulfate containing fragments from the heparan sulfates were compared by gel filtration and ion-exchange chromatography. The results showed that the 3T3 heparan sulfate contains 8% more O-sulfate than does the SV3T3 heparan sulfate. Analysis of uronic acids revealed that both types of heparan sulfates contain 45% L-iduronic acid and 55% D-glucuronic acid. These and other observations indicate that the primary difference in sulfation between the 3T3 and SV3T3 heparan sulfates lies in the extent of O-sulfation.     

Metabolism of sulfated glycosaminoglycans in cultivated bovine arterial cells. I. Characterization of different pools of sulfated glycosaminoglycans.

"Fibroblast-like" cells from the intimal layer of bovine aorta were grown in culture. The formation, composition, molecular weight and turnover rate of different pools of glycosaminoglycans were investigated in cultures incubated in the presence [35S]sulfate or [14C]glucosamine. The newly synthesized glycosaminoglycans are distributed into an extracellular pool (37 - 58%), a cell-membrane associated or pericellular pool (23 - 33%), and an intracellular pool (19 - 30%), each pool exhibiting a characteristic distribution pattern of chondroitin sulfate, dermatan sulfate, heparan sulfate and hyaluronate. The distribution pattern of the extracellular glycosaminoglycans resembles closely that found in bovine aorta. A small subfraction of the pericellular pool - tentatively named "undercellular" pool--has been characterized by its high heparan sulfate content. The intracellular and pericellular [35S]glycosaminoglycan pools reach a constant radioactivity after 8-12 h and 24 h, respectively, whereas the extracellular [35S]glycosaminoglycans are secreted into the medium at a linear rate over a period of at least 6 days. The intracellular glycosaminoglycans are mainly in the process of degradation, as indicated by their low molecular weight and by their half-life of 7 h, but intracellular dermatan sulfate is degraded more rapidly (half-life 4-5 h) than intracellular chondroitin sulfate and heparan sulfate (half-life 7-8 h). Glycosaminoglycans leave the pericellular pool with a half-life of 12-14 h by 2 different routes: about 60% disappear as macromolecules into the culture medium, and the remainder is pinocytosed and degraded to a large extent. Extracellular and at least a part of the pericellular glycosaminoglycans are proteoglycans. Even under dissociative conditions (4M guanidinium chloride) their hydrodynamic volume is sufficient for partial exclusion from Sepharose 4B gel. The existence of topographically distinct glycosaminoglycan pools with varying metabolic characteristics and differing accessibility for degradation requiresa reconsideration and a more reserved interpretation of results concerning the turnover rates of glycosaminoglycans as determined in arterial tissue.     

The binding and processing of plasminogen by Balb/c 3T3 and SV3T3 cells

The binding and processing of plasminogen by Balb/c 3T3 and SV3T3 cells was studied using ¹²⁵I-labeled canine plasminogen. Throughout a 3-day period, ¹²⁵I-plasminogen in the incubation medium bound to the cells and was degraded, first to intermediate-sized macromolecules that were the same size as the large (74,600-dalton) and small (25,000-dalton) chains of active plasmin, and to smaller fragments including 3-iodo-L-tyrosine. Binding to SV3T3 cells was independent of the protease-dependent morphological change (PDMC)¹ characteristic of these and many other transformed cells. The SV3T3, and to a somewhat lesser extent, the 3T3 cells, both accumulated and released into the incubation medium 3-iodo-L-tyrosine, a terminal lysozymal digestion product. The results of a sublethal cell-surface trypsinization assay suggest that the cell-associated plasminogen was primarily bound to the surfaces of the 3T3 and SV3T3 cells while the macromolecular degradation products including active plasmin were inside the cells. The rate of ¹²⁵I plasminogen degradation exhibited by SV3T3 cells was approximately two times greater than that of 3T3 cells, which presumably reflects differences in endocytosis or lysosomal hydrolysis, or both. The rates were unaffected by addition of pancreatic or soybean trypsin inhibitor sufficient to inhibit PDMC. In the incubation medium, plasminogen was activated to plasmin by SV3T3, but not by 3T3 cells. However, 95–100% of plasmin covalently bound to a 47,000-dalton canine serum component, which could be dissociated from plasmin by hydroxylamine: 95–100% of the plasmin was inactive to reaction with DF³²P. Thus the serum component is a plasmin inhibitor. The plasmin-containing complex in the medium had an apparent molecular weight of 212,000. Under denaturing conditions, the complex dissociated into two covalently modified plasmin-containing species of 153,000 and 127,000 daltons. In addition to forming a complex with a serum component, the plasmin is cleaved into two small fragments (～10,000 and 12,000 daltons) by as-yet uncharacterized serum factors.     

Effect of monensin on the sulfation of heparan sulfate proteoglycan from endothelial cells.

Monensin is a monovalent metal ionophore that affects the intracellular translocation of secretory proteins at the level of trans-Golgi cisternae. Exposure of endothelial cells to monensin results in the synthesis of heparan sulfate and chondroitin sulfate with a lower degree of sulfation. The inhibition is dose dependent and affects the ratio [35S]-sulfate/[3H]-hexosamine of heparan sulfate from both cells and medium, with no changes in their molecular weight. By the use of several degradative enzymes (heparitinases, glycuronidase, and sulfatases) the fine structure of the heparan sulfate synthesized by control and monensin-treated cells was investigated. The results have shown that among the six heparan sulfate disaccharides there is a specific decrease of the ones bearing a sulfate ester at the 6-position of the glucosamine moiety. All other biosynthetic steps were not affected by monensin. The results are indicative that monensin affects the hexosamine C-6 sulfation, and that this sterification is the last step of the heparan sulfate biosynthesis and should occur at the trans-Golgi compartment.     

The decreased synthesis of chondroitin sulfate-containing extracellular proteoglycans by SV40 transformed Balb/c 3T3 cells

Balb/c 3T3 cells synthesize 5–10 times more ³⁵SO₄^2?-labeled extracellular proteoglycan per cell than do Balb/c 3T3 cells transformed by SV40 (SV3T3). The extracellular ³⁵SO₄^2?-labeled proteoglycans of the Balb/c 3T3 and SV3T3 cells differ markedly in their acid mucopolysaccharide composition. Extracellular Balb/c 3T3 proteoglycans contain about 70–80% chondroitin sulfate, most of which is chondroitin 4-sulfate, and small amounts of heparan sulfate and/or heparin. On the other hand, extracellular SV3T3 proteoglycans contain 65–75% heparan sulfate and/or heparin and less than 15% chondroitin sulfate. Analysis of extracellular ³⁵SO₄^2?-labeled proteoglycan by sodium dodecyl sulfate-polyacrylamide gel electrophoresis reveals that Balb/c 3T3 alone synthesizes a class of proteoglycans capable of migrating in a 10% separating gel. This class of proteoglycans, designated as fraction C, accounts for up to 45% of the total extracellular Balb/c 3T3 ³⁵SO₄^2?-labeled proteoglycans and contains chondroitin sulfate exclusively. It is altogether absent in the extracellular SV3T3 proteoglycans. The absence of this and other classes of chondroitin sulfate-containing proteoglycans can account for the 5–10-fold decreased synthesis of ³⁵SO₄^2?-labeled proteoglycans by SV3T3 cells when compared to Balb/c 3T3 cells.     

Growth control in cultured 3T3 fibroblasts II. Molecular properties of a fraction enriched in growth inhibitory activity

Treatment of sparse, proliferating cultures of 3T3 cells with medium conditioned by exposure to density-inhibited 3T3 cultures resulted in an inhibition of growth and division in the target cells when compared to similar treatment with unconditioned medium. This growth inhibitory activity was fractionated by ammonium sulfate precipitation and gel filtration, yielding one fraction that was 35-fold enriched in specific activity. Analysis of the chemical and biological properties of this highly active fraction indicated that: (a) it is an endogenous cell product, synthesized by the 3T3 cells and shed into the medium; (b) it is a protein and its activity is sensitive to treatment with pronase; (c) the constituent polypeptide chains have molecular weights of 10,000 and 13,000; and (d) it is not cytotoxic and its effect on target cells are reversible. These results suggest that we have partially purified from conditioned medium an endogenous growth regulatory factor that may play a role in density-dependent inhibition of growth in cultured fibroblasts. We propose the term Fibroblast Growth Regulator to describe this class of molecules.     

Kinetics of proteoheparan sulfate synthesis, secretion, endocytosis, and catabolism by a hepatocyte cell line

The metabolism of heparan sulfate proteoglycan was studied in monolayer cultures of a rat hepatocyte cell line. Late log cells were labeled with 35SO4(2-) or [3H] glucosamine, and labeled heparan sulfate, measured as nitrous acid-susceptible product, was assayed in the culture medium, the pericellular matrix, and the intracellular pools. Heparan sulfate in the culture medium and the intracellular pools increased linearly with time, while that in the matrix reached a steady-state level after a 10-h labeling period. When pulse-labeled cells were incubated in unlabeled medium, a small fraction of the intracellular pool was released rapidly into the culture medium while the matrix heparan sulfate was taken up by the cells, and the resulting intracellular pool was rapidly catabolized. The structures of the heparan sulfate chains in the three pools were very similar. Both the culture medium pool and the cell-associated fraction of heparan sulfate contained proteoheparan sulfate plus a polydisperse mixture of heparan chains which were attached to little, if any, protein. Pulse-chase data suggested that the free heparan sulfate chains were formed as a result of catabolism of the proteoglycan. When NH4Cl, added to inhibit lysosomal function, was present during either a labeling period or a chase period, the total catabolism of the heparan sulfate chains to monosaccharides plus free SO2-4 was blocked, but the conversion of the proteoglycan to free heparan sulfate chains continued at a reduced rate.     

Relationship of transformation, cell density, and growth control to the cellular distribution of newly synthesized glycosaminoglycan

Mouse 3T3 cells and their Simian Virus 40-transformed derivatives (3T3SV) were used to assess the relationship of transfromation, cell density, and growth control to the cellular distribution of newly synthesized glycosaminoglycan (GAG). Glucosamine- and galactosamine- containing GAG were labeled equivalently by [3H=A1-glucose regardless of culture type, allowing incorporation into the various GAG to be compared under all conditions studied. Three components of each culture type were examined: the cells, which contain the bulk of newly synthesized GAG and are enriched in chondroitin sulfate and heparan sulfate; cell surface materials released by trypsin, which contain predominantly hyaluronic acid; and the media , which contain predominantly hyaluronic acid and undersulfated chondroitin sulfate. Increased cell density and viral transformation reduce incorporation into GAG relative to the incorporation into other polysaccharides. Transformation, however, does not substantially alter the type or distribution of newly synthesized GAG; the relative amounts and cellular distributions were very similar in 3T3 and 3T3SV cultures growing at similar rates at low densities. On the other hand, increased cell density as well as density-dependent growth inhibition modified the type and distribution of newly synthesized GAG. At high cell densities both cell types showed reduced incorporation into hyaluronate and an increase in cellular GAG due to enhanced labeling of chondroitin sulfate and heparan sulfate. These changes were more marked in confluent 3T3 cultures which also differed in showing substantially more GAG label in the medium and in chondroitin-6-sulfate and heparan sulfate at the cell surface. Since cell density and possibly density- dependent inhibition of growth but not viral transformation are major factors controlling the cellular distribution and type of newly synthesized GAG, differences due to GAG's in the culture behavior of normal and transformed cells may occur only at high cell density. The density-induced GAG alterations most likely involved are increased condroitin-6-sulfate and heparan sulfate and decreased hyaluronic acid at the cell surface.     

A transformation-dependent difference in the heparan sulfate associated with the cell surface.

Glycosaminoglycans from the surface of cultured mouse cells (3T3, SV40-3T3, 3T6) were released by trypsin digestion and separated by ion-exchange chromatography into hyaluronic acid, heparan sulfate and chondroitin sulfate. Using a double label technique, the glycosaminoglycans from 3T3 cells were compared with those from SV40-3T3 and 3T6 cells. No differences were apparent in either the hyaluronic acid or chondroitin sulfate fractions, however, the heparan sulfate from 3T3 cells was found to elute from DEAE-cellulose at a higher ionic strength than that from transformed cells. This altered behavior implies a structural difference in the cell surface heparan sulfate which appears to be dependent upon transformation.     

Density-dependent changes in the amount of sulfated glycosaminoglycans associated with mouse 3T3 cells.

The relative amount of sulfated glycosaminoglycans associated with the cell layer of parent and SV40-transformed Swiss mouse 3T3 cells was determined from the incorporation of labeled sulfate (35SO4) into macromolecular material. In cultures of SV40-transformed cells, the glycosaminoglycan content per cell was constant over a wide range of densities. In cultures of parent 3T3 cells, the glycosaminoglycan content per cell increased directly with density, the highest values being found in contact-inhibited cultures. At high cell densities, the glycosaminoglycan content of 3T3 cells was several-fold higher than that for SV40-transformed cells. Most of the density-dependent increase in glycosaminoglycans of 3T3 cells was accounted for by chondroitin sulfate (dermatan sulfate) which was over 6-fold higher in confluent cultures than in low density cultures.     

Transformed mouse mammary epithelial cells synthesize undersulfated basement membrane proteoglycan

Proteoglycans deposited in the basal lamina of [14C] glucosamine-labeled normal and [3H]glucosamine-labeled transformed mouse mammary epithelial cells grown on type I-collagen gels, were extracted in 4 M guanidinium chloride and cofractionated over Sepharose CL 4B. The heparan sulfate chains carried by these proteoglycans were isolated by treatment with alkaline borohydride, protease K, chondroitinase ABC, and cetylpyridinium chloride precipitation. Heparan sulfate isolated from transformed cell cultures consistently eluted from DEAE-cellulose at lower salt concentrations and was of smaller apparent Mr when chromatographed over Sepharose CL 6B, than heparan sulfate of normal cell cultures. Experiments using doubly labeled cultures ([3H]glucosamine and [35S]sulfate) demonstrated an approximately 30% reduction in the sulfate/hexosamine ratio in heparan sulfate derived from transformed cultures. Both N- and O-sulfate were decreased. The decreased Mr and decreased sulfation of heparan sulfate upon transformation appear sufficient to explain the altered heparan sulfate/chondroitin sulfate ratios previously observed in these cells. These changes may have implications for the molecular interactions in which these proteoglycans are normally engaged during basal lamina assembly, and cause the poor basal lamina formation displayed by these transformed cells.     

Regulation of haemopoiesis in long-term bone marrow cultures. IV. Glycosaminoglycan synthesis and the stimulation of haemopoiesis by beta- D-xylosides

Sulfated glycosaminoglycans (GAGs) are distributed in consistent and distinctive patterns between the cell surface and the growth medium of haemopoietically active long-term bone marrow cultures. Heparan sulfate is the main cell surface component and chondroitin sulfate is the major sulfated species in the medium. When the cultures are supplemented with beta-D-xylosides a significant increase in chondroitin sulfate synthesis is observed but no stimulation of heparan sulfate synthesis occurs. The chondroitin sulfate accumulates in the culture medium in beta-D-xyloside-treated cultures but the composition of sulfated GAGs in cell-surface derived material is unaffected. beta-D-xylosides also stimulate the production of haemopoietic cells without any apparent alteration in the adherent stromal cells of the marrow cultures. Equivalent increases are obtained in cells at all stages of development so that a fivefold increase in pluripotent stem cells (CFU-S) is matched by fivefold increase in the granulocyte-macrophage progenitors (GM-CFC) and in mature granulocytes. The stimulation persists for many weeks in beta-D-xyloside-treated cultures. These results indicate that the sulfated GAGs may play an important role in the regulation of haemopoiesis.     

A unique heparan sulfate in the nuclei of hepatocytes: structural changes with the growth state of the cells

Growing and confluent cultures of a rat hepatocyte cell line were labeled with 35SO4(2-) and the heparan sulfate in the culture medium, the pericellular matrix, the nucleus, the nuclear outer membrane, and the remaining cytoplasmic pool was purified by DEAE-cellulose chromatography. The heparan sulfate in all pools from the confluent cells was bound more strongly on the DEAE-cellulose column than the corresponding pools from the growing cells. Gel filtration of each pool before and after beta-elimination showed that the heparan sulfate from the nuclear and nuclear membrane pools was composed of primarily free chains, whereas the heparan sulfate in all of the other pools was a mixture of proteoglycans and free chains. The heparan sulfate in each pool was cleaved with nitrous acid to obtain mixtures of di- and tetrasaccharides. Analysis of these mixtures showed that the structural features of the heparan sulfates in each pool were different and were altered significantly when the growing cells became confluent. The nuclear-plus-nuclear membrane pools represented 6.5% and 5.4% of the total cell-associated heparan sulfate in the growing cells and the confluent cells, respectively. The structural features of the heparan sulfate in the two nuclear pools were very similar to each other, but were markedly different from those of the heparan sulfate from the other pools or from any previously described heparan sulfate or heparin. The most unusual aspect of these structures was the high content of beta-D-glucuronosyl(2-SO4)----D-glucosamine-N,O-(SO4)2 disaccharide units in these sequences. The mode of biosynthesis and delivery of these unusual sequences to the nucleus and the potential significance of these observations are discussed.